NEWOfivio POS is live — sell at the counter, track the stock, reorder before you run out.See what it does →
Design··3 min read

Percentage, lump sum or hourly — how to price architectural work

Most practices pick a fee model once and apply it forever. The model is a risk allocation, and the right allocation changes from project to project.

By Ofivio

There are three fee models in architecture and every practice has a favourite. That is the problem: a fee model is not a house style, it is a decision about who carries which risk, and the right answer changes with the project.

Percentage of construction cost

The traditional instrument, and still the most defensible on a large project with an uncertain scope. The fee scales with the thing being designed, which is broadly fair: a bigger building is more work.

  • **Rewards** — a scope that grows. If the client adds a floor, the fee follows without a renegotiation.
  • **Punishes** — a client who value-engineers. You do the work of designing the expensive version and are paid on the cheap one.
  • **Watch** — what "construction cost" means. Does it include the client's direct contracts, the fit-out, the external works? Define it in the appointment or you will define it in an argument.

Lump sum

A fixed fee for a defined scope. Clients like it because it is a number; practices like it until the scope moves.

  • **Rewards** — efficiency. If you deliver it in less time than you priced, that is your margin.
  • **Punishes** — a client who changes their mind, unless variations are written into the appointment and actually claimed. Most practices write them in and never claim them.
  • **Watch** — the number of design iterations included. Three is a common figure. Without it, "just try one more option" is unbounded.

Hourly

The honest instrument for genuinely open-ended work — feasibility, advisory, expert opinion, a project whose scope nobody can describe yet.

  • **Rewards** — nothing, and that is the point. You are paid for effort, so nobody is gambling.
  • **Punishes** — a practice with no reliable record of its own time, which is most of them.
  • **Watch** — clients read an hourly rate as an invitation to audit. If you bill hourly you must be able to produce a defensible breakdown without a week's reconstruction.

Choosing per project

If the projectUseBecause
Has a scope you can describe in a pageLump sumYou can price it, and efficiency is yours to keep
Has a budget but no defined scopePercentageThe fee moves with the thing, so neither side is guessing
Has neitherHourly, with a cap and a reviewNobody can price uncertainty; do not pretend otherwise
Is a repeat client on a known typeLump sumYou have done it before; that experience is worth money
Includes construction supervisionSplit the feeDesign and supervision are different risks — see below
The question is who is better placed to carry the uncertainty.

The split nobody applies

Where the practice also supervises construction, the design fee and the supervision fee are two different instruments with two different risk profiles, and lumping them into one percentage is how supervision ends up unbilled when a project stalls on site. Design fee vs supervision fee works through it.

Whatever you choose, you need to know what it cost you

Every model above is a bet, and you cannot tell whether you are winning bets without knowing what a project cost in staff time. That is not an argument for timesheets everywhere — see should architects use timesheets — but it is an argument for knowing, at minimum, who was on which project and for roughly how long.

Ofivio holds the fee, the stages and the staff cost on one project, so the question is answered rather than reconstructed.

See how it works
Filed under
architect fees·pricing·practice management·fee models